Commodity vs Non-Commodity: The London Local Content Google Wants

There is a sentence in Google's guide to generative AI search that I think is the most important thing it says for local businesses, and it is not about AI at all. It says creating content people find unique, compelling and useful will likely influence your presence in generative AI search in the long run more than any of the other suggestions in the guide.

Most of the guide is about what to ignore. This is the part about what to do, and it is the part I find people nod at and then ignore. So here is a practical version for a London local business, built from Google's own language.

What is commodity content?

Commodity content is content based on common knowledge that could originate from anyone and typically adds little unique insight. Google's example is '7 Tips for First-Time Homebuyers'. Its counter-example, non-commodity content, is 'Why We Waived the Inspection and Saved Money: A Look Inside the Sewer Line', which gives an experienced take that goes beyond the ordinary.

The word commodity is exactly right. A commodity is interchangeable: one barrel of oil is much like another. A commodity page is interchangeable in the same way. If you could swap the businesses' names and nobody would notice, there is no reason for a search system to choose yours, and no reason for an AI system to quote it.

Why does Google care about this now?

Because AI systems read many sources, so a page that restates what the others say adds nothing to an answer built from all of them. Google puts it directly: its AI systems look at a variety of sources, so a unique viewpoint stands out, and a summary of existing content simply restates information already available elsewhere.

That is a structural change in what makes a page valuable. In classic search, a page could win by being a better-optimised version of the common answer. In a system that retrieves from many pages and writes a synthesis, the common answer is already in the synthesis. What you can contribute is the part nobody else has.

What are the attributes of non-commodity content?

Google lists a unique point of view, helpful and reliable people-first writing, clear organisation and supporting images or video. It asks you to create the content yourself based on what you know and to consider what in-depth experience you can bring.

• A unique point of view. Google's example is a first-hand review versus a summary of existing content. For a local firm, that is a named person describing work they actually did.
• Helpful, reliable, people-first writing. Written for a reader who wants to decide something, not for a crawler.
• Organised for readers. Paragraphs, sections and headings that make the page easy to scan and navigate.
• High-quality images and video. Google says its generative AI features can bring in images and video, which creates more ways to appear beyond a page link. Use your own photographs, not stock.

What does non-commodity content look like for a London local business?

It is usually a page with an experience in it: a job, a number, a decision, a mistake or a result that only the business could report. I find five formats work reliably, and none of them need a big budget.

1. The worked job. One real piece of work, what was found, what was done, what it cost in time, what the customer said. Get written permission before naming anyone.
2. The local fact sheet. Practical details for an area: parking rules, access, building types, typical problems. A plumber's page about pre-war housing stock in a particular part of London is something only a local can write.
3. The honest comparison. Two ways of doing a thing and which you would choose and why, including when the cheaper one is the right answer.
4. The data you hold. Search Console data, call logs and grid scans you can legitimately publish. My case studies do this, including the numbers that are not flattering.
5. The correction. A page saying 'this is the common advice, here is why it is wrong in our experience', with evidence.

Notice that every one of those requires access to something real. That is the point. You cannot generate access to a real job, and that is why this kind of content is defensible.

Can AI tools help me write it?

Yes, as a drafting and editing aid, provided the content still meets Search Essentials and the spam policies. Google says as much, and warns against recycling what others have said or what could easily be produced by a generative AI model.

My rule is that the model never supplies the facts. I put in the notes from the job, the numbers, the photographs and my opinion, and use the tool to tidy and structure. If you ask a model to write a page about boiler servicing in Camden with no inputs, it returns something true of everywhere. That is a commodity page produced at speed. Google's guidance on core updates points the same way: pages made for search engines first and people second are the ones at risk.

How do I publish real jobs without exposing a client?

Get written permission before you name anyone, and anonymise anything personal by default. First-hand content is only an advantage if you can publish it safely, and in the UK identifiable details about a customer, such as their name, address, photographs of their home or health information, are personal data you should not publish without a proper basis.

• Ask at the point of the job. A simple permission line on the sign-off form is far easier than chasing consent months later.
• Strip identifying detail unless the client has agreed to be named: street numbers, faces, number plates, documents in shot.
• Describe the problem, not the person. 'A 1930s semi in north-west London with a failed damp course' is specific and safe. A name and a house number is neither.
• Never invent. A composite job presented as a real one is a fabricated case study, which is worse than having no case study.
• Keep the evidence. Photos, dates and job notes should be on file, so that if anyone asks you can show the work happened.

This is the same standard I hold the case studies on this site to: a named source for every number, permission before naming a client, and the unflattering figures left in. It makes the pages slower to produce and much harder to copy, which is exactly the property non-commodity content is supposed to have.

How do I test whether a page is commodity?

Run the swap test, then the source test. If a direct competitor could replace your name with theirs and publish the page unchanged, it is commodity. If every sentence could have been written without ever doing the work, it is commodity.

What should I do with the content I already have?

1. List your ten highest-value pages by enquiries, not traffic.
2. Run the swap test on each. Mark the ones that fail.
3. Add one first-hand element to each failing page. A job, a photograph, a number, a named person. Do not rewrite the whole page.
4. Leave the pages that already pass alone. Google advises avoiding changes to content that is performing well.
5. Plan one new non-commodity piece a month. A steady rhythm of real material beats a burst of volume.

This is slow, unglamorous and entirely within your control. It is also, in my reading of Google's own words, the single most durable thing a local business can do. Everything else in the AI conversation is either a technical hygiene point or a hack Google has told you to ignore.